Measuring environmental exposures in people's activity space: The need to account for travel modes and exposure decay

Lai Wei*, Mei-Po Kwan, Roel Vermeulen, Marco Helbich

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Background: Accurately quantifying people’s out-of-home environmental exposure is important for identifying disease risk factors. Several activity space-based exposure assessments exist, possibly leading to different exposure estimates, and have neither considered individual travel modes nor exposure-related distance decay effects. Objective: We aimed (1) to develop an activity space-based exposure assessment approach that included travel modes and exposure-related distance decay effects and (2) to compare the size of such spaces and the exposure estimates derived from them across typically used activity space operationalizations. Methods: We used 7-day-long global positioning system (GPS)-enabled smartphone-based tracking data of 269 Dutch adults. People’s GPS trajectory points were classified into passive and active travel modes. Exposure-related distance decay effects were modeled through linear, exponential, and Gaussian decay functions. We performed cross-comparisons on these three functional decay models and an unweighted model in conjunction with four activity space models (i.e., home-based buffers, minimum convex polygons, two standard deviational ellipses, and time-weighted GPS-based buffers). We applied non-parametric Kruskal–Wallis tests, pair-wise Wilcoxon signed-rank tests, and Spearman correlations to assess mean differences in the extent of the activity spaces and correlations across exposures to particulate matter (PM2.5), noise, green space, and blue space. Results: Participants spent, on average, 42% of their daily life out-of-home. We observed that including travel modes into activity space delineation resulted in significantly more compact activity spaces. Exposure estimates for PM2.5 and blue space were significantly (p < 0.05) different between exposure estimates that did or did not account for travel modes, unlike noise and green space, for which differences did not reach significance. While the inclusion of distance decay effects significantly affected noise and green space exposure assessments, the decay functions applied appear not to have had any impact on the results. We found that residential exposure estimates appear appropriate for use as proxy values for the overall amount of PM2.5 exposure in people’s daily lives, while GPS-based assessments are suitable for noise, green space, and blue space. Significance: For some exposures, the tested activity space definitions, although significantly correlated, exhibited differing exposure estimate results based on inclusion or exclusion of travel modes or distance decay effect. Results only supported using home-based buffer values as proxies for individuals’ daily short-term PM2.5 exposure.

Original languageEnglish
Number of pages9
JournalJournal of Exposure Science and Environmental Epidemiology
Volume33
Issue number6
Early online date14 Feb 2023
DOIs
Publication statusPublished - 2023

Keywords

  • Daily mobility
  • Environmental health
  • Exposome
  • Exposure assessment
  • GPS tracking
  • Individual exposure

Fingerprint

Dive into the research topics of 'Measuring environmental exposures in people's activity space: The need to account for travel modes and exposure decay'. Together they form a unique fingerprint.

Cite this